Carbon Fiber Tensile Strength vs. Shear Strength: Which Property Matters Most for Your Application? (Engineering Guide)

An Engineering Guide

By  Zhao Yuan Zheng, CEO & Technical Director, Xiangsheng Composite Materials Co., Ltd.

Last technically reviewed: June 2026, based on in-house tensile, shear, and torsional testing on carbon fiber tubes, boards, and woven laminates.

QUICK ANSWER Tensile strength tells you how a carbon fiber part behaves when you pull it in a straight line, along the fiber direction. Shear strength tells you how it behaves when layers slide against each other, when it twists, or when it is loaded through a bolted or bonded joint. The two numbers come from different parts of the material: fibers carry tension, while resin and the fiber-matrix bond carry shear. One is rarely a reliable predictor of the other. If your application involves torque, joints, or short, thick sections, shear strength deserves more attention than most datasheets give it.

 

Why I Keep Getting Asked This Question

Over the past four years running the technical side of Xiangsheng, I’ve answered some version of this question more times than I can count. It usually arrives by email, often at an odd hour, because the buyer is somewhere in Europe or North America and it’s already evening on their end. The pattern is almost always the same: an engineer has a datasheet in front of them with one number circled, almost always tensile strength, and wants to know whether that number alone tells them if the part will hold up. My answer is almost always that it depends on how the part actually fails. That’s not what someone wants to hear when they’re trying to close out a procurement spec by Friday afternoon, but it’s the honest one.

I’m writing this guide because I’d rather spend twenty minutes explaining the difference now than spend three months later helping a client work out why a part that tested beautifully in a straight-pull test failed within weeks once it went into torsion service. That second scenario has happened to clients of ours more than once before they came to us, and it’s a large part of the reason this article exists.

Tensile Strength and Shear Strength Come From Two Different Parts of the Material

Before getting into which property matters for which application, it helps to understand why these two numbers can diverge so dramatically on the very same laminate. They aren’t two measurements of how strong the material is in some general sense. They are measurements of two entirely different load-carrying mechanisms inside the composite.

What Tensile Strength Actually Measures

When you pull a unidirectional carbon fiber laminate along the fiber axis, the load travels almost entirely through the carbon fibers themselves. The resin matrix plays a supporting role: it keeps the fibers aligned, spaces them evenly, and transfers small amounts of load between adjacent fibers, but the fiber itself is doing the heavy lifting. That’s why tensile strength, tested under ASTM D3039 or ISO 527-4/5, scales so closely with fiber type, fiber volume fraction, and how precisely the fibers stay aligned with the load direction. A standard-modulus carbon/epoxy laminate with good fiber alignment and a typical 55-60% fiber volume fraction will usually test somewhere in the 1,500 to 2,500 MPa range along the 0° direction, and that number tends to stay fairly stable across different resin systems, simply because the resin isn’t the limiting factor in this load case.

What Shear Strength Actually Measures

Shear strength is a different story, and there isn’t just one version of it. Interlaminar shear strength, tested by the short beam method under ASTM D2344 or ISO 14130, measures how well adjacent plies resist sliding past one another. In-plane shear strength, usually measured through a tensile test on a ±45° laminate under ASTM D3518, or through the V-notched rail method in ASTM D7078, measures resistance to shear within a single ply orientation. Torsional shear capacity in a tube is its own animal again, governed by winding angle and wall architecture rather than by a single coupon-level test.

What all three have in common is that they depend far more on the resin and the fiber-matrix interface than on the fiber itself. Resin toughness, the quality of the fiber sizing chemistry, void content left over from the cure cycle, and the stacking sequence of the plies all move the needle on shear strength in ways that barely touch tensile strength at all. I’ve personally seen two laminates with nearly identical 0° tensile numbers post interlaminar shear values that differ by 30 percent or more, simply because one used a tougher resin formulation and had a cleaner autoclave cure with lower void content.

If you’re used to specifying steel or aluminum, this divergence can be counterintuitive. Both metals are close to isotropic, so their shear strength runs at a fairly predictable 50 to 60 percent of tensile strength in any direction, because the metal’s internal structure doesn’t care which way it’s loaded. Carbon fiber composites don’t get that consistency. Because the fiber and resin play such different roles, the same laminate can show a shear-to-tensile ratio many times lower than a metal would, and the exact ratio depends entirely on fiber orientation. A 0° tensile number that would tell you nearly everything about a steel bracket tells you comparatively little about how the same-shaped composite part will hold up under torque or at a joint.

How Far Apart Are the Numbers, Really?

To make this concrete, here is how these properties typically compare on a standard carbon/epoxy system, based on our own in-house testing and cross-referenced against published composite materials data.

Property	Governing Test Standard	Typical Range (Carbon/Epoxy, ~55-60% Fiber Volume)	What Mainly Controls It
0° Tensile Strength	ASTM D3039 / ISO 527-5	1,500 – 2,500 MPa	Fiber type, fiber volume fraction, fiber alignment
Interlaminar Shear Strength (ILSS)	ASTM D2344 / ISO 14130	60 – 100 MPa	Resin toughness, fiber-matrix bonding, void content
In-Plane Shear Strength	ASTM D3518 / D7078	70 – 120 MPa	Resin shear response, ply stacking, fiber/matrix interface
Torsional Shear Capacity (Tube)	Custom torque-to-failure testing	Highly layup-dependent	Winding angle (±45° content), wall thickness, mandrel quality

 

The gap shown above is really the entire point of this article. A laminate rated at 2,000 MPa in tension might only be delivering 70 to 90 MPa in interlaminar shear. That isn’t a manufacturing defect. It’s simply how fiber-reinforced composites are built. The mistake I see buyers make most often is treating a high tensile number as evidence of an all-around “strong” material, when really it’s only evidence that the material is strong in one specific direction, under one specific kind of load.

The Real Question Isn’t Which Property Is Higher, It’s Which One Your Load Path Touches

I’d rather buyers stop asking “which property matters more in general” and start asking “which property governs the failure mode my part is actually exposed to.” Here’s how I walk clients through that distinction.

Pure Tension or Compression Members

Tie rods, tension cables, straight structural struts, and anything where load travels in a straight line along the part’s main axis are tensile-dominated applications. Fiber alignment matters enormously here; even a few degrees of misalignment from the load direction can cost you a meaningful percentage of rated tensile strength. For these parts, I tell clients to spend their engineering attention on fiber straightness and 0° content, because tensile strength is genuinely the number to watch.

Torque and Torsion: Where Shear Quietly Takes Over

Drive shafts, robotic arm links, transmission tubes, and anything that twists rather than stretches are shear-dominated, full stop. This is one of the most common misunderstandings I run into with carbon fiber tube buyers, especially those coming from a metals background, where torque capacity and tensile strength tend to move together more predictably. In a composite tube, a layup optimized purely for axial tensile and bending performance, heavy on 0° fiber, will often deliver disappointing torque capacity, because 0° fibers are nearly useless for resisting shear. It’s the ±45° plies that carry torsional load in a filament-wound or rolled tube, and if a tube’s layup is light on ±45° content, no amount of tensile strength printed on the datasheet will save it from a twisting failure.

 

0° fiber carries a straight pull. ±45° fiber is what actually resists a twist.

Bending and the Span-to-Thickness Ratio

Flexural loading is where I see the most confusion, because bending genuinely does involve tension on the outer fiber of the tension face and compression on the opposite face, but it also generates shear stress through the thickness of the part, concentrated near the neutral axis. Whether a part fails by tensile or compressive fiber failure, or by interlaminar shear instead, depends heavily on the span-to-thickness ratio of the section. This is exactly why composite flexural test standards such as ASTM D7264 typically specify span-to-thickness ratios in the range of 16:1 to 32:1: the ratio is chosen specifically to keep the failure mode tensile and compressive rather than shear-dominated, so the test reports a clean flexural strength number. Shrink that ratio in a real part, a short, thick bracket or a stub cantilever, and shear stress at the mid-plane can become the limiting factor well before the outer fiber ever reaches its tensile capacity.

 

Same load, same materials. Geometry alone decides whether the part fails in tension or in shear.

Bolted and Bonded Joints: Where Tensile Rating Almost Never Matters

This is the one I probably explain most often. A bolted composite joint typically fails by bearing failure or shear-out around the hole, not by net tensile failure of the laminate cross-section. A bonded joint fails at the adhesive bond line, governed by bond-line shear strength, not by the tensile strength of either substrate. I’ve reviewed more than a few specifications over the years where the buyer sized a joint purely off the laminate’s published tensile strength, with no bearing or shear-out calculation at all. That’s a gap I’d encourage every composite buyer to close before finalizing a joint design.

Specifying by the Wrong Property: Pros, Cons, and Risk

Buyers usually default to one of three approaches when writing a purchase spec, and each carries a different level of risk depending on what the part actually has to survive.

Specification Approach	Pros	Cons	Risk Level
Specify by tensile strength only	Simple, single number; easy to compare across suppliers; sufficient for true axial tension members	Says nothing about torque, joint, or bending-shear performance, the most common cause of the field failures we see	High for torsion, joints, or short/thick parts; low for pure tension members
Specify by shear data only	Directly captures the matrix- and joint-governed failure modes that actually break tubes and brackets	Doesn’t confirm axial load capacity; can under-spec a part that’s genuinely tension-dominated	Low for torsion/joint parts; medium for tension members if axial isn’t also checked
Specify both, mapped to the load case	Matches the property requested to the failure mode the part will see; lets a supplier optimize 0° vs ±45° content instead of guessing	Requires the buyer to communicate the real load case up front, which takes slightly more time at the RFQ stage	Low across all load types when done correctly

 

In our experience, the small amount of extra work in the third row is what separates parts that perform consistently in the field from parts that pass the datasheet check and then fail once they meet a real load case.

Two Cases I Still Reference With My Engineering Team

Two projects come up often enough in conversations with new clients that I’ll walk through both here.

Case One: Torsional Splitting in a Robotic Arm Link

A few years ago, a robotics integrator in Europe came to us after a frustrating experience with a previous supplier. They were using carbon fiber tubes as rotating links in an automated arm, subject to repeated torque reversals during normal operation. The tubes had performed exactly as promised on paper, with a strong axial tensile rating and good bending stiffness, but in the field, several units developed longitudinal splitting along the tube wall within a matter of weeks. That’s a classic torsional shear failure, not a tensile one.

When we pulled the original datasheet, the problem was visible almost immediately to anyone who knew where to look. The layup was heavily 0°-dominant, clearly optimized for axial strength and bending stiffness, with only a thin ±45° content. It had never actually been torque-tested; the torque rating on the original spec sheet had simply been calculated from the tensile properties, a shortcut that works fine for some materials and fails badly for filament-wound composite tubes.

Our team ran torque-to-failure testing alongside the standard tensile and interlaminar shear coupons, then redesigned the layup with a calculated ±45°/0°/±45° stacking sequence sized specifically for the torque-reversal profile the client described. The redesigned tube increased torque-to-failure capacity by roughly two and a half times over the original, with about a 6 percent increase in wall mass, and held its axial tensile and bending performance within the client’s original specification. It has been running in their production line since, without a repeat of the splitting failures.

 

Same tube diameter, same client load case. The layup, not the tensile rating, decided whether it survived.

Case Two: A Bonded Bracket Joint That Looked Fine on Paper

A second pattern shows up just as often, this time with a bonded joint rather than a tube. A client building outdoor solar panel mounting brackets sourced a carbon fiber board bonded into an aluminum frame. The boards passed flexural and tensile testing with a comfortable margin, but several assemblies failed in the field in a way that looked, at first glance, like a tensile crack near the edge of the board.

When we inspected the failed parts, the crack ran cleanly along the adhesive bond line rather than through the laminate itself, which is the signature of a bond-line shear failure, not a tensile failure of the board. The original specification had never included bond-line shear data; the joint had been sized using the laminate’s flexural strength alone, with no test representing the actual bonded geometry.

We ran a single-lap shear test representative of the real joint, adjusted surface preparation and adhesive selection, and confirmed the redesigned joint against the client’s actual load case before it went back into production. After implementing the revised bonding process, field failures dropped from approximately 4% of installed assemblies during the first year to well below 0.3%, while the expected service life increased from around three years to more than eight years under the same operating conditions.

I bring these cases up with new clients fairly often, not to criticize the previous suppliers; mistakes like this happen across the industry. I bring them up because together they’re the clearest illustration I have of why asking for “the strength rating” without specifying which kind of strength is asking the wrong question.

How We Test and Specify Both Properties at Xiangsheng

Our lab runs tensile testing under ASTM D3039 and ISO 527, short beam interlaminar shear under ASTM D2344 and ISO 14130, and in-plane shear via ASTM D3518 and D7078, alongside custom torque-to-failure rigs for our carbon fiber tube products. For carbon fiber boards, our highest-volume product line and the one most of our overseas clients know us for first, we typically pair flexural testing with interlaminar shear data whenever a board will be bolted, clamped, or used inside a sandwich panel structure, since those joint conditions put shear strength front and center even though boards are most often specified by flexural or tensile properties first.

Because we operate our own high-pressure autoclave curing, prepreg lines, and weaving and filament-winding equipment in-house, we can adjust fiber orientation, ply count, and resin system on a layup-by-layup basis rather than offering one fixed product specification. When a client tells us their actual load case, torque-reversal profile, joint type, expected span and thickness, we design the layup around that and then test it to confirm it, instead of starting from a generic “high-strength” laminate and hoping it happens to fit.

A Practical Checklist Before You Specify Carbon Fiber for Your Application

A few habits I’d recommend to any engineer or buyer working with carbon fiber tubes, boards, or custom laminates for the first time:

• Identify the actual failure mode first. Decide whether your part is primarily loaded in tension, torsion, bending, or through a joint before you ask for a single “strength” number.
• Request both tensile and shear data. Ask for whichever values are relevant to your load case, not just whichever number the supplier leads with on the brochure.
• For tubes under torque, ask for torque-to-failure data. A torsional shear rating inferred from axial tensile strength isn’t the same as a tube that has actually been twisted to failure on a rig.
• For thick laminates or short brackets, check the span-to-thickness ratio. Ask what ratio was used in any flexural test you’re being shown, since a short, thick test article will report different numbers than a long, slender one.
• For bolted or bonded assemblies, request bearing and bond-line shear data. The tensile strength of the laminate itself is rarely the limiting factor at a joint.
• Match fiber orientation to load direction. 0° plies for straight pull, ±45° plies for torque and shear, and quasi-isotropic layups when the load direction is unpredictable or multi-axial.
• Ask which test standard backs every number on the datasheet. If a supplier can’t tell you the standard behind a strength value, treat that number with real caution.

 

Diagnosing Shear and Torsion Failures After the Fact

Not every project starts with a clean specification. If a part is already failing in the field, these are the patterns we see most often, along with how to confirm the cause and what to do about it.

Field Symptom	Likely Root Cause	How to Confirm	Recommended Fix
Tube splits or delaminates lengthwise under repeated twisting, despite passing tensile spec	Layup is 0°-dominant; torque rating was calculated from tensile data rather than tested	Check whether the torque number on the datasheet is calculated or tested; review the layup schedule for ±45° content	Rebalance toward a ±45°/0°/±45°-style stacking sequence sized to the torque-reversal profile, then verify with an actual torque-to-failure test
Bolted joint cracks or shows elongated holes well below the laminate’s rated tensile strength	Bearing or shear-out failure at the hole, not net-section tensile failure	Inspect the failure location; bearing and shear-out failures center on the bolt hole rather than mid-span	Increase edge distance, add a local doubler ply around the hole, or request bearing and shear-out data for your specific hole geometry
Short, thick bracket fails in a way that looks tensile despite a correctly specified resin and fiber	Span-to-thickness ratio too low; the true failure was interlaminar shear at the mid-plane	Compare the ratio used in the supporting flexural test (commonly 16:1 to 32:1 under ASTM D7264) against the part’s actual geometry	Re-test flexural performance at the part’s real ratio, or request dedicated interlaminar shear data instead
Two laminates with near-identical tensile numbers perform very differently under cyclic or joint loads	Resin toughness, fiber-matrix bonding, or cure void content differ between the two, none of which shows up in a tensile number	Request interlaminar shear and void-content data for both; compare cure process and parameters	Specify a minimum ILSS value and a maximum void content alongside tensile strength in the purchase spec

 

Quick Reference: Tensile vs. Shear at a Glance

For a fast gut-check before you dig into full engineering calculations:

Tensile Strength Matters Most When…	Shear Strength Matters Most When…
The load is a straight axial pull or tension member	The part is under torque, twisting, or torsional load
It’s a long, slender beam with a high span-to-thickness ratio	It’s a short, thick beam or bracket with a low span-to-thickness ratio
The load path is fiber-dominated, along the 0° direction	The part connects through a bolted, bonded, or clamped joint
Typical parts: tie rods, tethers, straight structural struts	Typical parts: drive shafts, robotic links, rotating tubes

 

Frequently Asked Questions

Is carbon fiber stronger in tension or in shear?

There is no universal answer; it depends on the laminate and the direction of loading. But as a general pattern, carbon fiber composites are typically far stronger in tension along the fiber direction than they are in shear. It’s common to see 0° tensile strength values 15 to 25 times higher than interlaminar shear strength on the very same laminate, simply because tension is fiber-dominated and shear is matrix-dominated.

How does carbon fiber’s tensile-to-shear gap compare to steel or aluminum?

It’s a much wider gap than most metals show. Steel and aluminum are close to isotropic, so their shear strength typically runs at roughly 50 to 60 percent of tensile strength regardless of loading direction. Carbon fiber composites don’t follow that pattern, because tension and shear are carried by different parts of the material. Depending on layup, shear strength can fall to a small fraction of the 0° tensile number, which is why a tensile rating that tells you almost everything about a steel part tells you comparatively little about how a composite part will hold up in torsion or at a joint.

What test standard measures carbon fiber shear strength?

There isn’t a single standard; it depends on which type of shear you need. Interlaminar shear, between plies, is most commonly tested by the short beam method under ASTM D2344 or ISO 14130. In-plane shear is typically measured via a tensile test on a ±45° laminate under ASTM D3518, or through the V-notched rail shear method in ASTM D7078. Tubular products under torsion are usually evaluated with a dedicated torque-to-failure test rather than a coupon-level standard.

Why did my carbon fiber tube perform well in a tensile test but fail under torque in the field?

This almost always traces back to fiber orientation. Tensile and bending performance are driven primarily by 0° fiber content, while torsional shear performance is driven primarily by ±45° fiber content. A tube optimized for axial strength can be genuinely weak in torsion if its layup doesn’t include enough fiber oriented to resist shear, which is exactly the failure pattern in the case study described above.

Does a higher tensile strength automatically mean higher shear strength?

No, and this is one of the most persistent misconceptions I encounter. Because the two properties are governed by different load-carrying mechanisms, it’s entirely possible, and common, for two laminates with similar tensile strength to show meaningfully different shear strength, depending on resin toughness, fiber-matrix bonding, void content, and layup sequence.

Can shear strength be improved without sacrificing tensile strength?

To a meaningful degree, yes. Adding or rebalancing ±45° plies, selecting a tougher resin system, and tightening void content during cure can all raise shear strength with only a modest effect on 0° tensile properties. There is always some trade-off in weight or stiffness, but in our experience, a well-engineered layup adjustment rarely requires giving up significant tensile performance to gain meaningful shear capacity.

What should I tell my supplier when requesting a custom carbon fiber tube or board?

At minimum, describe the actual load case: is the part primarily in tension, torsion, bending, or carrying load through a bolted or bonded joint? Include the expected load magnitude, the operating environment, and whether the load is static or cyclic. That information lets a supplier design a layup around your real failure mode, instead of defaulting to a generic high-tensile specification that may not match how your part will actually be loaded.

Final Thoughts From My Side of the Table

After four years of fielding these questions, the pattern I keep coming back to is this: tensile strength gets all the attention because it’s the easiest number to test, to print on a datasheet, and to compare across suppliers at a glance. Shear strength is messier; there are multiple kinds of it, the test methods are less standardized across the industry, and it doesn’t make for as clean a marketing number. None of that makes it less important. If anything, the failures I’ve personally helped diagnose over the years have almost always traced back to a mismatch between the property that was specified and the property that actually governed how the part failed.

My honest recommendation to anyone sourcing carbon fiber tubes, boards, or custom laminates for an overseas application is simple: ask your supplier for both numbers, ask which test standard backs each one, and make sure someone has actually mapped your load case to the right property before the part goes into production. It’s a five-minute conversation that can save months of field failures. If you’re still deciding between carbon fiber and another reinforcement option for the application altogether, our guide to comparing carbon fiber, fiberglass, aramid, and hybrid composites is a useful next stop.

NEED HELP SPECIFYING TENSILE OR SHEAR PERFORMANCE FOR A CUSTOM PROJECT? Our technical team at Xiangsheng Composite Materials can review your load case, run tensile, interlaminar shear, in-plane shear, or torque-to-failure testing in-house, and recommend a fiber layup matched to how your part will actually be loaded, whether that’s a custom carbon fiber tube, board, or woven laminate. Get in touch with our engineering team to discuss your application.

 

ABOUT THE AUTHOR Zhao Yuanzheng is CEO and Technical Director of Xiangsheng Composite Materials Co., Ltd., a China-based manufacturer of carbon fiber tubes, boards, woven fabric, and prepreg systems supplying technical composite components to clients in more than 30 countries. He leads an in-house technical team of more than ten engineers focused on mechanical testing, custom layup design, and production scale-up for tension-, torsion-, and joint-critical composite applications.